There are numerous definitions for sudden deaths. All of the definitions include two main parameters. They relate to the time from the onset of symptoms and signs, and the time of death.¹ These times vary between zero, 1, 6 and 24 h.¹ The WHO defines a sudden death as a death occurring within 24 hours after the onset of symptoms.² Byard classified these deaths into three categories: those individuals who were apparently completely well and died suddenly; those who were mildly unwell; and those with known serious disease processes but were stable on treatment.¹ The admission of cases of sudden, unexpected and/or unexplained deaths (SUU-Deaths) to medico-legal mortuaries is a well-known worldwide phenomenon, and in many of these cases an inadequate history is present at the time of the investigation.³

In South Africa, the performance of medico-legal autopsies is regulated by the Inquests Act (No. 58 of 1959). The definition of unnatural deaths is in the Regulations Regarding the Rendering of Forensic Pathology Service (R636) in accordance with Chapter 11 of the National Health Act (No. 61 of 2003). These include cases ‘where the death is sudden and unexpected, or unexplained, or where the cause of death is not apparent’.

SUU-Deaths are considered to be some of the most challenging autopsies that can be conducted by pathologists.³ In many of these cases, it is difficult to ascertain whether or not the person died from a disease process found at autopsy or as a result thereof.⁴

HIV can cause death of an individual in numerous different ways including AIDS, accelerated cardiovascular disease, malignancies and opportunistic infections.⁵ AIDS is regarded as being the number one cause of death in HIV patients.⁶ Tseng et al. indicated that sudden cardiac deaths in HIV-positive individuals accounted for 86% of all cardiac deaths with a mean rate of 2.6 per 1000 person-years.⁶

Antiretroviral therapies have been associated, in a number of studies, with the increased risk of cardiovascular disease.^7,8 It was documented that: ‘Increased exposure to protease inhibitors is associated with an increased risk of myocardial infarction, which is partly explained by dyslipidaemia’.⁹ TB and HIV co-infection may also be linked to SUU-Death. TB and HIV co-infected patients have a five times greater risk of dying within 2 years of receiving TB treatment.¹⁰ Other studies have determined that 31% of all new TB cases in the African region can be attributed to HIV infection: ‘TB was the cause of 11% of all adult AIDS-related deaths’. In South Africa, it was estimated that there were 2 million co-infected adults in 2003.^11,12

South Africa has a jaded past with misinformation and the stigmatisation of HIV-positive individuals.^13,14 Due to these social factors, the disclosure of an accurate medical history is not forthcoming in cases of SUU-Death. This ends up hindering the medico-legal investigation of the death.

South Africa had a population of 49 320 500 people as of mid-2009, and 10.6% of the population was HIV-positive.¹⁵ The burden of the disease on the state is large and costly. In order to minimise the consequences of the pandemic, every possible chance to understand and refine the information, used to make important prevention and treatment decisions, needs to be implemented and assessed.

This study seeks to provide relevant statistical information on the HIV prevalence rate in the SUU-Death population of the Pretoria Medico-Legal Laboratory (MLL). The results of this study could also provide additional insight into the epidemiology of the virus and the SUU-Death population. This can facilitate a protocol change in the handling of the SUU-Death population admitted to the Pretoria MLL. This would be to implement standardised HIV testing in SUU-Death cases, which, in turn, will have budgetary and policy implications for the Forensic Pathology Service (FPS).

The data collection took approximately 10 months to complete in 2009. A total of 130 samples were collected, of which 100 samples yielded a valid HIV result. The remaining 30 samples when tested provided an invalid result on the test (the absence of a test result). This was discussed in a previous study, and it could be caused by an increased post-mortem interval.¹⁷

The HIV prevalence in the SUU-Death population at the Pretoria MLL is 43%. This is significantly higher than the HIV prevalence in the general mortuary population at the Pretoria MLL (26%), translating into a 17% difference in prevalence between the two groups.¹⁷ This was determined to be statistically significant. White South Africans account for 18% of the general mortuary population and their HIV prevalence was 8% in the general mortuary population.¹⁷ However, in the SUU-Death population they accounted for an unexpected 19% of the population and had an HIV prevalence of 16%. Black South Africans accounted for 76% of the general mortuary population with an HIV prevalence of 32%.¹⁷ In the SUU-Death population, black South Africans accounted for 79% of the population but had an HIV prevalence of 51%. The concerning factor is that there are such large numbers of HIV-positive cases being admitted to the Pretoria MLL and that HIV is not being routinely tested for in the mortuary setting.

In contrast to the study by Christiansen in which he found the top three causes of SUU-Death – cardiovascular, malignancy and infection, the majority of our cases died from respiratory disease processes followed by cardiovascular disease. TB accounted for 12 of the 31 respiratory disease-related deaths. In South Africa’s Millennium Development Goals, Goal 6 declares that there is a co-infection rate of 70% with TB and HIV.²¹ In a report by Statistics South Africa in 2009, TB accounted for 12% of all deaths, influenza and/or pneumonia 7.5% and HIV 3.1%,²² thus giving us a context for this difference in profile.

TB–HIV co-infection and antiretroviral treatment side effects are two possible explanations that warrant further investigation in this subpopulation. Tseng et al. indicated that sudden cardiac deaths in HIV-positive individuals accounted for 86% of all cardiac deaths with a mean rate of 2.6 per 1 000 person-years.⁶ In our study, cardiovascular diseases accounted for 17% of the cases with the majority being due to cardiomyopathy.

SUU-Deaths are a challenging autopsy and need to be approached with caution and the right amount of scepticism. The purpose of these autopsies is firstly to rule out an unnatural cause of death and secondly to provide, if possible, the cause of death even if it is natural.

The value of post-mortem HIV testing as an ancillary investigative technique to assist in the decision-making process and diagnosis of the cause of death should be further investigated. These cases are often admitted with no available history as the next of kin is not known or the medical history is incomplete. An HIV test can be helpful in determining the cause of death.

Unfortunately in South Africa, HIV has not been routinely tested for in the SUU-Death cases, and as a result statistics and observations regarding this challenging autopsy population have left a number of unanswered questions. This research study has attempted to address these questions and has shown that the analysis and observations provided by this testing provide an accurate well-documented source of information to facilitate the complex decisions that need to be taken in determining the cause of death and ultimately in our fight with regard to the HIV pandemic.

If medical records, post-mortem HIV test results and eyewitness testimony are made available before autopsy, a death scene investigator (DSI) can collect the relevant records and information and present them to the forensic pathologist along with the HIV test results. This would simplify the post-mortem examination for the forensic pathologist and facilitate a possible reduction in the number of SUU-Death cases admitted to the Pretoria MLL for a full autopsy. A reduction in SUU-Death cases could save the FPS a considerable amount of money and decrease the workload. This will ultimately streamline the provision of FPSs in the Pretoria MLL. A spin-off of testing for HIV routinely would be that there is a database of results that could be used to better understand HIV in the PM setting.

SUU-Deaths accounted for 14% of all cases admitted to the Pretoria MLL. The HIV prevalence in the SUU deaths was 43%, which is 17% higher than the general mortuary population in Pretoria (p = 0.0045). The majority of these deaths were due to respiratory disease processes with 12 cases having HIV–TB co-infection.

This study demonstrates the value of testing the SUU-Death population for HIV at autopsy. Further research recommendations are that research should be conducted in the implementation of routine HIV testing and testing for which stage of infection the HIV-positive SUU-Death cases are in. Screening for resistance and antiretrovirals could also be undertaken to determine if the person was on treatment at the time of death.